Mar Biol (2011) 158:2695–2703 DOI 10.1007/s00227-011-1767-7 ORIGINAL PAPER Functional effects of the hadal sea cucumber Elpidia atakama (Echinodermata: Holothuroidea, Elasipodida) reflect small-scale patterns of resource availability A. J. Jamieson • A. Gebruk • T. Fujii • M. Solan Received: 19 May 2011 / Accepted: 27 July 2011 / Published online: 7 August 2011 Ó Springer-Verlag 2011 Abstract Holothuroidea represent the dominant benthic terms of abundance and biomass (Rice et al. 1982; Ohta megafauna in hadal trenches (*6,000–11,000 m), but little 1983; Sibuet 1985; Billett 1991). A consistent feature of is known about their behaviour and functional role at such holothurian communities, irrespective of location, is the depths. Using a time-lapse camera at 8,074 m in the Peru– marked increase in diversity at abyssal depths (3,000– Chile Trench (SE Pacific Ocean), we provide the first in 6,000 m) (Billett 1991) relative to bathyal (1,000–3,000 m) situ observations of locomotory activity for the elasipodid (Hansen 1975) and hadal depths ([6,000 m) (Hansen 1957; holothurian Elpidia atakama Belyaev in Shirshov Inst Belyaev 1989). Frequently observed mass abundances of Oceanol 92:326–367, (1971). Time-lapse sequences reveal holothurians, particularly in trenches associated with high ‘run and mill’ behaviour whereby bouts of feeding activity productivity in temperate and sub-Antarctic latitudes, have are interspersed by periods of locomotion. Over the total led some authors to refer to the hadal zone as ‘‘the kingdom of observation period (20 h 25 min), we observed a mean Holothuroidea’’ (sensu Belyaev 1989), a view that has been (±SD) locomotion speed of 7.0 ± 5.7 BL h-1, but this reinforced by trawl-catch frequencies of 88% at depths increased to 10.9 ± 7.2 BL h-1 during active relocation [6,000 m (comparable only to Polychaeta) and high levels and reduced to 4.8 ± 2.9 BL h-1 during feeding. These of dominance (75–98% in number of all organisms retrieved, observations show E. atakama translocates and processes [90% biomass) at depths [7,500 m (Belyaev 1989). sediment at rates comparable to shallower species despite Although such high returns tend to be associated with trawls extreme hydrostatic pressure and remoteness from surface- confined to the bottom of the axial part of the trenches, where derived food. the greatest quantity of organic matter accumulates (Otosaka and Noriki 2000; Danovaro et al. 2003; De Leo et al. 2010; Jamieson et al. 2010), it follows that the cumulative contri- Introduction bution of the Holothuroidea to deep ocean benthic process and functioning must be considerable (Amaro et al. 2010). Deposit feeding invertebrates, such as the Holothuroidea, Despite such high levels of abundance, intra- and inter- dominate benthic megafaunal communities in the deep sea in specific competition is thought to be low because indi- vidual species of holothurians adopt different feeding strategies, including preferential feeding on nutritionally Communicated by S. Uthicke. rich patches (Hauksson 1979; Hudson et al. 2005) and/or subtle differences in mobility or feeding behaviour & A. J. Jamieson ( ) Á T. Fujii Á M. Solan (Hudson et al. 2005; Godbold et al. 2009) that allow them Oceanlab, Institute of Biological and Environmental Sciences, University of Aberdeen, Main Street, Newburgh, to utilise different fractions of the same detrital food Aberdeenshire AB41 6AA, UK source (Uthicke and Karez 1999; Miller et al. 2000). It is e-mail: [email protected] also known that differences in feeding rates (compare, for example, Hudson et al. 2005; Godbold et al. 2009) alter A. Gebruk P.P. Shirshov Institute of Oceanology, Russian Academy the gut residence time of a food parcel, leading to more of Sciences, Nakhimovsky Pr. 36, Moscow 117997, Russia efficient digestion and rates of assimilation (Hiratsuka and 123 2696 Mar Biol (2011) 158:2695–2703 Uehara 2007). Whilst these physiological and behavioural (Echevin et al. 2008), a region of high surface productivity adaptations influence holothurian activity, much of the (averaging 269 g C m-2 year-1, Longhurst et al. 1995) ingested material is of low nutritional value (Lopez and with values reported as high as 3,613 g C m-2 year-1 Levinton 1987), leading to foraging activity that results in (Fossing et al. 1995). significant levels of surficial bioturbation that has seldom been quantified (Sibuet and Lawrence 1981; Bett and Rice 1993; Uthicke 1999; Roberts et al. 2000; Bett et al. 2001). Equipment Indeed, a recent review of invertebrate bioturbation (Teal et al. 2008) indicates a paucity of such data from bathyal We deployed Hadal-lander B (Jamieson et al. 2009b), a or abyssal depths (maximum depth data obtained = free-falling lander equipped with a 5 megapixel digital still 5,654 m, Yang et al. 1986) and a complete absence of camera (OE14-208; Kongsberg Maritime, UK) and a information from hadal depths. Hence, it is clear that the Conductivity, Temperature and Depth (CTD) sensor (SBE- ecological consequences of particle redistribution follow- 19plus V2; SeaBird Electronic Inc. USA), at 8,074 m in the 0 ing holothurian foraging and feeding activities are not Richards Deep, Peru–Chile trench (23° 22.470 S, 71° 0 known despite the importance of this group in deep ocean 19.973 W) on the 13 September 2010. The camera was mounted vertically (altitude 1 m) providing a visible area ecosystems. -2 In contrast to shallower environments, where species of 62 9 46.5 cm (0.29 m ). We attached a bait (*1kg can be caught and returned to the laboratory, carrying out of Tuna, Thunnus sp.) to a 1-cm-diameter scaled bar in the controlled experimental manipulations on hadal specimens centre of the field of view (FOV) and positioned to inter- is difficult. Direct in situ experimental manipulations are sect the sediment–water interface. Time-lapse images were possible using remotely operated vehicles (ROVs), but taken at 60-s intervals. The CTD probe recorded temper- progress is slow as there is only one vehicle capable of ature (°C), salinity, and pressure (dbar) every 10 s. Lander sampling [6,000 m (Fletcher et al. 2010). A more prac- recover was achieved by acoustically jettisoning ballast tical solution is the use of free-fall baited cameras weights to initiate the ascent to the surface. CTD data were (Jamieson et al. 2009a, b) which allow time-lapse or averaged, and pressure was converted to depth (m) continuous video recordings of benthic faunal behaviour. following Saunders (1981). Here, we use such technology to present the first detailed in situ account of the locomotion and feeding behaviour of Image analysis the holothurian Elpidia atakama (Belyaev 1971) (family Elpidiidae) at 8,074 m in the Richards Deep area of the Motion paths of E. atakama were tracked using ImageJ Peru–Chile trench. This species has never been seen alive 1.42q, a Java-based public domain program developed at and appears to be endemic to the Peru–Chile trench. We the USA National Institutes of Health (available at, compare locomotion speed and feeding behaviour with http://rsb.info.nih.gov/ij/index.html). Images were cali- abyssal analogues and conclude that the observed patterns brated and analysed in chronological order. We used the of activity are consistent with the view that the behaviour X–Y coordinates at the base of the central feeding tentacle of E. atakama reflects exploitation of patchily distributed as a location marker and calculated the distance (cm) an resources. We contend that the ability to exploit envi- individual travelled per time step to determine the absolute locomotion speed (cm h-1) as follows: ronmental heterogeneity in this way explains, at least in qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi part, why the holothurians are so abundant and outcom- 2 2 d ðÞX2 À X1 þ ðÞY2 À Y1 pete other deposit feeding species at abyssal and hadal a ¼ ¼ ð1Þ depths. t t where a˜ = absolute locomotion speed, cm h-1; d = dis- tance, cm; and t = time interval (here, 60 s). We also explore how differences in time-lapse interval Materials and methods alter ecological interpretation by re-calculating the absolute locomotion speed (and percentage deviation in error rela- Study site tive to the highest resolution of observation) for E. atakama at 1-, 2-, 5-, 10-, 20- and 60-min intervals and place our The Peru–Chile Trench (SE Pacific Ocean) is the longest results within the context of other findings published in the trench in the world (5,900 km 9 100 km, Angel 1982) and scientific literature. runs parallel to the west coast of South America from In order to account for body size, we divided a˜ by body Ecuador to central Chile. The trench lies below the length (BL = longest axis of specimen) to provide a size- Humboldt Current and the Peruvian upwelling system specific speed (BL h-1). 123 Mar Biol (2011) 158:2695–2703 2697 Results these together, a˜ = 35.5 ± 29.3 cm h-1 (7.0 ± 5.7 BL h-1), giving a mean swept area rate (a˜ 9 body width 9 We recorded 1,225 images (20 h 25 min) at a calculated distance/t) of 81.6 cm-2 h-1. The speed and direction of depth of 8,072 m (=8,276 dbar). At this depth, water movement was neither constant nor unidirectional (Fig. 2), temperature was 2.25°C and salinity was 34.68. Sequences reflecting a ‘run and mill’ movement pattern (Kaufmann included observations of the holothurian, Elpidia atakama, and Smith 1997), i.e. relatively large distances are achieved and three species of scavenging gammarid amphipods: with minimal changes in speed and direction (routine Eurythenes gryllus, (Thurston et al. 2002), Hirondellea locomotion) and are interspersed with bouts of localised, sp.nov.